Search results should be unbiased. Social media platforms should be neutral. The Internet should be for everyone.
From Facebook to TikTok to DuckDuckGo, tech companies are facing pressure to take explicit stands against Russia’s ongoing invasion of Ukraine. With some exceptions, tech has responded to the call, at the cost of its relationships with Russia.
But there’s another cost to doing what many see as the right thing in Ukraine. It requires the tech companies to acknowledge in a very public way that their products and policies aren’t neutral after all — and it reminds us all of their own unchecked power over the world’s information systems.
Just two months ago, the privacy-focused search engine DuckDuckGo was trending on Twitter because conservative pundits and activists were hailing it as an unbiased alternative to Google. Specifically, they appreciated that DuckDuckGo surfaced results that espoused a conspiracy theory pushed by a recent guest of the podcaster Joe Rogan, whereas Google’s results were mostly debunkings.
They may not have realized that DuckDuckGo’s results, which are supplied in large part by Microsoft’s Bing, weren’t “unbiased” on purpose. They were merely less effective than Google’s at surfacing more reliable sources of information. DuckDuckGo knew that, of course. But over the years, it had encouraged such misunderstandings by promoting its search engine as “neutral” and “unbiased.”
On Friday, DuckDuckGo found itself trending among conservatives again, but this time it was on the wrong end of their ire. The reason: DuckDuckGo CEO and founder Gabriel Weinberg had announced that the search engine would begin down-ranking sites associated with Russian disinformation.
“This is not the way bro,” one libertarian influencer shot back on Twitter, garnering more than 22,000 likes. “We no longer trust anyone to decide for us what is ‘misinformation’. Let us make our own calls about that. Otherwise you’re just another tentacle for some Ministry of Truth.”
The hashtag this time: #DuckDuckGone.
The reality, as all major search engine providers have long privately understood, is that there is no such thing as an unbiased search algorithm. The whole point of search engines is to rank results according to some subjective set of criteria, which typically include things like the relevance of the page, the reputation of the source, and feedback from users and testers about the quality of the results. Surfacing reliable information over disinformation has always been the goal.
The difference is that it’s usually done in secret, via tweaks to the code, which may prove more effective for some search queries than others. By acknowledging publicly that it would specifically down-rank Russian disinformation, DuckDuckGo was bringing the quiet part into the open. And the reaction from the right served as a reminder of why many tech companies prefer to avoid doing that whenever possible.
Facebook’s race-blind practices around hate speech came at the expense of Black users, new documents show
Facebook, meanwhile, has spent years trying to thread a needle between presenting itself as a neutral platform for all ideas on the one hand, and a responsible actor that polices hate, misinformation, and threats to the integrity of democratic elections on the other. All the while, it has relied heavily on a lengthy, byzantine set of “community standards” that it purports to apply objectively, without political favor, in deciding which posts and accounts to allow and which to take down.
Of course, there is no perfect, global rule book for the bounds of acceptable speech. And neutrality, even if it were possible, is a dubious north star, one that leads to false equivalence between truthtellers and liars, or oppressors and the oppressed. And so, in practice, Facebook is constantly bending, amending, and adding to those rules as circumstances and public pressures necessitate.
But rarely has the company publicly embraced a double standard as explicitly as it did on Thursday, when it said it would make exceptions to its rules against violent and hateful speech for users in Ukraine and some neighboring countries. For instance, it will no longer take down a post from a Ukrainian who says “death to the Russian invaders.” That comes a week after the company said it would change its rules to allow people to praise neo-Nazis in Ukraine in the context of their resistance to the Russian invasion.
It’s an understandable move: to ban or suppress the speech of Ukrainian users advocating or organizing resistance at a time when their country is under attack and their lives are on the line would seem cruel, maybe even evil. Yet it raises fair questions about why Facebook has refused to take such a principled stand in other contexts, such as its long-standing “race-blind” policies that equated criticism of White people with anti-Black racism.
TikTok, even more than Facebook, has taken pains to present itself as apolitical — a place for fun and entertainment. But no platform of that scale can truly be apolitical, and the war in Ukraine has made it clearer than ever that TikTok is a battleground of information and ideas.
On Friday, The Washington Post reported that the White House had arranged a briefing on Ukraine for 30 TikTok stars. Meanwhile, Vice News reported that other TikTok stars were being paid to toe the Kremlin’s party line.
Early on in the conflict, TikTok followed Facebook and YouTube in banning Russian state media in Europe, partly in response to pressure from Ukrainian leaders. And last weekend, it suspended the posting of new videos from Russia in response to a Russian “fake news” law.
Scholars who study the impacts of technology on society will tell you that tech is virtually never neutral, and almost no algorithm is unbiased. What we should demand of technology companies, instead of neutrality, are transparency as to the biases built into their products, and accountability for their effects. It’s unfortunate that it has taken a world-shaking war to jolt them in that direction.